Monitoring effects on IAT Final exercises

Findings

• The monitoring intervention increased the likelihood of the student skipping the IAT score section.
• Students in the monitored groups also had higher IAT scores.
• The monitoring intervention in the later course rounds increased likelihood of student saying their professors provided motivation during the course.
• Students in monitored classes reported less agreement with the statement “There is at least one topic or activity that I never get bored thinking about”.

Effect on exit exercise decisions

outcome_cols = c('started_exercise' = 'Started exercise', 
                    'completed_survey' = "Completed survey",
                    'redistribute_decision' = "Redistribute decision",
                    'dictator_decision' = "Dictator decision",
                    'lying_payout' = "Lying payout",
                    'trolley_decision' = "Trolley decision",
                    'skip_iat' = "Skip IAT score",
                    'iat_score' = "IAT Score",
                    'book_econ' = "Economics book",
                    'book_gender' = "Gender book",
                    'book_morality' = "Morality book",
                    'book_independence' = "Judicial independence book")

note <-  "Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level."

# models_plot <- feols(fml = c(started_exercise, completed_survey, redistribute_decision,
#                              dictator_decision, lying_payout, trolley_decision, 
#                              skip_iat,iat_score, book_econ, book_gender, book_morality,
#                              book_independence) ~ i(round, monitoreo) | idcurso,
#                      data = data,
#                      split = ~round, cluster = ~idcurso+dni)

for (outcome in names(outcome_cols)) {
  
  formula <- as.formula(paste0(outcome, ' ~ monitoreo * teacher_female | idcurso | 0 | aula'))
  
  out_list <- list()
  
  for (r in 1:9) {
    model <- felm(formula, data = data %>% filter(round == r))
    
    out_list[[sprintf('Round %i', r)]] <- model
  }
  
  formula <- as.formula(paste0(outcome, ' ~ monitoreo * teacher_female | 0 | 0 | dni'))
  
  out_list[['Pure Treatment']] <- felm(formula, data = pure_control)
  out_list[['Perc. courses monitored']] <- felm(formula, data = monit_perc)
  
  
  out_name <- outcome_cols[[outcome]]
  cat('<h4></h4>')
  cat(sprintf('<h3>%s</h3>', out_name))
  
  stargazer(out_list,
            covariate.labels = c('Monitoring', 'Female Teacher', 'Monitoring x Female Teacher'),
            dep.var.labels = out_name,
            column.labels = names(out_list),
            notes = note,
            omit = 'Constant',
            type = 'html')
  
  feols(fml = as.formula(sprintf('%s ~ i(round, monitoreo) | idcurso', outcome)),
                       data = data, cluster = ~aula) %>%
    iplot(dict = outcome_cols, xlab = 'Round') 

}

Started exercise

	Dependent variable:
	Started exercise
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.033	-0.051	0.059	0.014	-0.009	-0.004	0.002	-0.004	-0.069	-0.026	-0.002
	(0.042)	(0.067)	(0.040)	(0.056)	(0.051)	(0.053)	(0.054)	(0.053)	(0.079)	(0.041)	(0.037)
Female Teacher	-0.289***	-0.044	0.084**	0.033	-0.008	-0.058*	-0.048	0.135***	-0.100*	0.011	-0.006
	(0.031)	(0.041)	(0.042)	(0.059)	(0.059)	(0.035)	(0.034)	(0.046)	(0.054)	(0.027)	(0.022)
Monitoring x Female Teacher		0.105	-0.160	-0.056	0.057		-0.009		0.121	-0.00000	-0.003
	(0.000)	(0.099)	(0.150)	(0.106)	(0.117)	(0.000)	(0.124)	(0.000)	(0.088)	(0.038)	(0.034)
Observations	558	558	558	557	536	534	534	529	532	3,550	4,845
R2	0.023	0.004	0.009	0.002	0.002	0.012	0.015	0.017	0.008	0.001	0.0001
Adjusted R2	0.018	-0.003	0.002	-0.006	-0.006	0.005	0.005	0.009	0.001	0.0003	-0.001
Residual Std. Error	0.376 (df = 554)	0.380 (df = 553)	0.379 (df = 553)	0.381 (df = 552)	0.385 (df = 531)	0.382 (df = 529)	0.382 (df = 528)	0.382 (df = 524)	0.383 (df = 527)	0.393 (df = 3546)	0.381 (df = 4841)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

Completed survey

	Dependent variable:
	Completed survey
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.012	-0.090	-0.003	-0.052	-0.072	-0.065	-0.043	-0.053	-0.115	-0.071	-0.063
	(0.043)	(0.063)	(0.042)	(0.063)	(0.058)	(0.054)	(0.055)	(0.053)	(0.092)	(0.044)	(0.041)
Female Teacher	-0.355***	-0.095**	0.061	-0.013	-0.017	-0.082**	-0.102***	0.059	-0.120**	-0.008	-0.027
	(0.029)	(0.042)	(0.049)	(0.075)	(0.066)	(0.038)	(0.036)	(0.044)	(0.053)	(0.030)	(0.025)
Monitoring x Female Teacher		0.154	-0.156	-0.004	0.095		-0.023		0.131	0.010	0.005
	(0.000)	(0.108)	(0.180)	(0.102)	(0.116)	(0.000)	(0.132)	(0.000)	(0.095)	(0.042)	(0.039)
Observations	558	558	558	557	536	534	534	529	532	3,550	4,845
R2	0.030	0.008	0.009	0.004	0.007	0.007	0.015	0.005	0.015	0.007	0.005
Adjusted R2	0.024	0.001	0.002	-0.003	-0.001	-0.0002	0.006	-0.003	0.008	0.006	0.004
Residual Std. Error	0.424 (df = 554)	0.429 (df = 553)	0.429 (df = 553)	0.431 (df = 552)	0.433 (df = 531)	0.431 (df = 529)	0.431 (df = 528)	0.432 (df = 524)	0.430 (df = 527)	0.428 (df = 3546)	0.429 (df = 4841)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

Redistribute decision

	Dependent variable:
	Redistribute decision
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.173	0.092	0.015	0.292	0.362	0.331	0.465**	0.319	0.194	0.419**	0.395**
	(0.190)	(0.248)	(0.266)	(0.225)	(0.283)	(0.209)	(0.182)	(0.221)	(0.326)	(0.198)	(0.184)
Female Teacher	0.281	0.206	-0.253	0.362	0.319	-0.384**	0.905***	-0.156	-0.240	0.206	0.134
	(0.272)	(0.375)	(0.335)	(0.365)	(0.363)	(0.171)	(0.164)	(0.127)	(0.221)	(0.139)	(0.122)
Monitoring x Female Teacher		0.091	0.561	-0.250	-0.175		-0.951***		0.152	-0.169	-0.108
	(0.000)	(0.450)	(0.357)	(0.390)	(0.412)	(0.000)	(0.214)	(0.000)	(0.358)	(0.186)	(0.175)
Observations	456	456	456	455	435	434	434	429	432	2,847	3,948
R2	0.004	0.007	0.008	0.008	0.011	0.022	0.029	0.018	0.005	0.013	0.010
Adjusted R2	-0.003	-0.002	-0.001	-0.001	0.001	0.013	0.018	0.008	-0.005	0.012	0.009
Residual Std. Error	1.777 (df = 452)	1.783 (df = 451)	1.782 (df = 451)	1.777 (df = 450)	1.783 (df = 430)	1.781 (df = 429)	1.769 (df = 428)	1.774 (df = 424)	1.793 (df = 427)	1.760 (df = 2843)	1.773 (df = 3944)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

Dictator decision

	Dependent variable:
	Dictator decision
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.119	0.307	-0.092	0.194	0.337**	0.187	0.069	0.062	0.118	0.101	0.106
	(0.173)	(0.243)	(0.229)	(0.141)	(0.141)	(0.137)	(0.137)	(0.151)	(0.173)	(0.203)	(0.189)
Female Teacher	0.173	0.426	-0.309	0.346	0.509**	0.271**	0.706***	-0.037	-0.144	0.170	0.033
	(0.118)	(0.419)	(0.209)	(0.358)	(0.234)	(0.124)	(0.096)	(0.139)	(0.215)	(0.152)	(0.131)
Monitoring x Female Teacher		-0.601	0.568**	-0.299	-0.617*		-0.414***		-0.110	-0.090	0.046
	(0.000)	(0.489)	(0.251)	(0.433)	(0.353)	(0.000)	(0.148)	(0.000)	(0.227)	(0.201)	(0.188)
Observations	454	454	454	453	433	432	432	427	430	2,829	3,930
R2	0.004	0.008	0.008	0.006	0.010	0.004	0.009	0.0004	0.004	0.001	0.001
Adjusted R2	-0.003	-0.001	-0.001	-0.003	0.001	-0.005	-0.003	-0.009	-0.005	0.0004	0.0002
Residual Std. Error	1.839 (df = 450)	1.842 (df = 449)	1.843 (df = 449)	1.843 (df = 448)	1.876 (df = 428)	1.880 (df = 427)	1.877 (df = 426)	1.863 (df = 422)	1.884 (df = 425)	1.807 (df = 2825)	1.855 (df = 3926)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

Lying payout

	Dependent variable:
	Lying payout
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.0002	0.015	0.022	0.031**	-0.003	0.025	0.018	0.027	0.034	0.024	0.019
	(0.025)	(0.037)	(0.022)	(0.016)	(0.033)	(0.028)	(0.029)	(0.029)	(0.022)	(0.032)	(0.028)
Female Teacher	0.138***	0.013	0.032	0.111**	-0.020	-0.010	-0.011	-0.038	0.117	0.029	0.021
	(0.020)	(0.032)	(0.035)	(0.054)	(0.040)	(0.024)	(0.025)	(0.023)	(0.074)	(0.026)	(0.018)
Monitoring x Female Teacher		-0.007	0.019	-0.069	0.052		0.045		-0.114	-0.003	0.003
	(0.000)	(0.052)	(0.062)	(0.077)	(0.064)	(0.000)	(0.057)	(0.000)	(0.080)	(0.035)	(0.031)
Observations	452	452	452	451	431	430	430	425	428	2,811	3,912
R2	0.009	0.002	0.005	0.016	0.004	0.003	0.004	0.004	0.013	0.002	0.002
Adjusted R2	0.002	-0.007	-0.004	0.007	-0.006	-0.006	-0.007	-0.005	0.004	0.001	0.001
Residual Std. Error	0.285 (df = 448)	0.286 (df = 447)	0.286 (df = 447)	0.285 (df = 446)	0.293 (df = 426)	0.293 (df = 425)	0.294 (df = 424)	0.295 (df = 420)	0.293 (df = 423)	0.338 (df = 2807)	0.291 (df = 3908)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

Trolley decision

	Dependent variable:
	Trolley decision
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.024	0.021	0.004	0.024	0.047	-0.015	0.005	-0.030	0.011	0.007	0.008
	(0.038)	(0.042)	(0.040)	(0.048)	(0.056)	(0.039)	(0.043)	(0.037)	(0.044)	(0.049)	(0.045)
Female Teacher	-0.083*	-0.010	-0.016	-0.030	0.047	-0.059	0.026	-0.040*	-0.105	-0.008	-0.011
	(0.049)	(0.071)	(0.065)	(0.066)	(0.060)	(0.040)	(0.040)	(0.021)	(0.091)	(0.032)	(0.026)
Monitoring x Female Teacher		0.006	0.053	-0.017	-0.087		-0.098		0.021	-0.016	-0.019
	(0.000)	(0.082)	(0.094)	(0.069)	(0.062)	(0.000)	(0.060)	(0.000)	(0.096)	(0.044)	(0.040)
Observations	451	451	451	450	430	429	429	424	427	2,802	3,903
R2	0.003	0.002	0.002	0.003	0.003	0.008	0.010	0.011	0.009	0.0003	0.001
Adjusted R2	-0.004	-0.007	-0.007	-0.006	-0.006	-0.002	-0.001	0.001	-0.001	-0.001	-0.0003
Residual Std. Error	0.420 (df = 447)	0.420 (df = 446)	0.420 (df = 446)	0.420 (df = 445)	0.423 (df = 425)	0.421 (df = 424)	0.421 (df = 423)	0.422 (df = 419)	0.421 (df = 422)	0.423 (df = 2798)	0.420 (df = 3899)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

Skip IAT score

	Dependent variable:
	Skip IAT score
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.087***	0.076	0.055*	0.060***	0.054**	0.092***	0.091***	0.079**	0.084**	0.117**	0.101**
	(0.029)	(0.050)	(0.031)	(0.021)	(0.026)	(0.031)	(0.034)	(0.032)	(0.034)	(0.048)	(0.044)
Female Teacher	0.013	-0.004	-0.052	-0.060	-0.039	-0.059**	0.086***	0.142***	0.011	0.001	-0.003
	(0.024)	(0.044)	(0.033)	(0.054)	(0.049)	(0.024)	(0.023)	(0.028)	(0.058)	(0.031)	(0.026)
Monitoring x Female Teacher		0.017	0.088**	0.122	0.123*		0.033		-0.013	0.035	0.039
	(0.000)	(0.067)	(0.042)	(0.077)	(0.068)	(0.000)	(0.048)	(0.000)	(0.068)	(0.045)	(0.042)
Observations	461	461	461	460	440	439	439	434	437	2,873	3,991
R2	0.012	0.011	0.015	0.017	0.020	0.017	0.030	0.023	0.011	0.022	0.015
Adjusted R2	0.006	0.002	0.007	0.009	0.011	0.008	0.018	0.014	0.001	0.021	0.014
Residual Std. Error	0.411 (df = 457)	0.412 (df = 456)	0.411 (df = 456)	0.411 (df = 455)	0.408 (df = 435)	0.409 (df = 434)	0.405 (df = 433)	0.408 (df = 429)	0.413 (df = 432)	0.420 (df = 2869)	0.408 (df = 3987)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

IAT Score

	Dependent variable:
	IAT Score
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.069**	0.101***	0.084*	0.099***	0.071*	0.078**	0.081**	0.097**	0.037	0.094**	0.089**
	(0.033)	(0.036)	(0.046)	(0.035)	(0.041)	(0.032)	(0.032)	(0.038)	(0.039)	(0.044)	(0.041)
Female Teacher	0.006	0.009	0.020	-0.006	-0.062	-0.056**	-0.182***	-0.016	-0.083***	-0.042	-0.034
	(0.047)	(0.053)	(0.052)	(0.047)	(0.050)	(0.025)	(0.017)	(0.034)	(0.027)	(0.031)	(0.025)
Monitoring x Female Teacher		-0.073	-0.025	-0.126**	-0.003		0.178***		0.024	-0.005	-0.001
	(0.000)	(0.077)	(0.068)	(0.060)	(0.062)	(0.000)	(0.035)	(0.000)	(0.045)	(0.040)	(0.037)
Observations	292	292	292	291	277	278	278	274	275	1,754	2,530
R2	0.014	0.018	0.015	0.030	0.024	0.019	0.032	0.027	0.011	0.025	0.018
Adjusted R2	0.004	0.005	0.002	0.016	0.009	0.005	0.014	0.012	-0.003	0.024	0.017
Residual Std. Error	0.307 (df = 288)	0.306 (df = 287)	0.306 (df = 287)	0.305 (df = 286)	0.306 (df = 272)	0.307 (df = 273)	0.305 (df = 272)	0.307 (df = 269)	0.306 (df = 270)	0.301 (df = 1750)	0.304 (df = 2526)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

Economics book

	Dependent variable:
	Economics book
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.0002	0.005	-0.0003	0.005	0.043***	0.012	0.017	0.002	-0.036*	-0.008	-0.00004
	(0.017)	(0.031)	(0.021)	(0.015)	(0.017)	(0.021)	(0.022)	(0.021)	(0.020)	(0.028)	(0.027)
Female Teacher	0.012	0.006	0.005	-0.017	0.028	0.082***	0.049***	-0.078***	-0.075***	-0.001	-0.002
	(0.020)	(0.024)	(0.022)	(0.027)	(0.018)	(0.017)	(0.018)	(0.015)	(0.025)	(0.017)	(0.014)
Monitoring x Female Teacher		0.010	0.022	-0.012	-0.113***		-0.080*		0.058*	-0.002	-0.011
	(0.000)	(0.048)	(0.026)	(0.050)	(0.022)	(0.000)	(0.045)	(0.000)	(0.034)	(0.021)	(0.021)
Observations	439	439	439	438	419	418	418	413	416	2,721	3,801
R2	0.001	0.002	0.002	0.002	0.013	0.004	0.003	0.006	0.010	0.0004	0.0002
Adjusted R2	-0.006	-0.007	-0.008	-0.007	0.004	-0.006	-0.009	-0.004	0.0003	-0.001	-0.001
Residual Std. Error	0.265 (df = 435)	0.265 (df = 434)	0.265 (df = 434)	0.265 (df = 433)	0.254 (df = 414)	0.255 (df = 413)	0.256 (df = 412)	0.252 (df = 408)	0.255 (df = 411)	0.242 (df = 2717)	0.259 (df = 3797)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

Gender book

	Dependent variable:
	Gender book
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.042	0.061	0.068	0.022	0.042	0.034	0.055	0.042	0.014	0.037	0.032
	(0.042)	(0.069)	(0.054)	(0.050)	(0.055)	(0.050)	(0.049)	(0.043)	(0.041)	(0.058)	(0.054)
Female Teacher	-0.190***	-0.020	0.045	-0.040	-0.011	0.123***	-0.112***	-0.067*	0.022	-0.043	-0.019
	(0.030)	(0.086)	(0.052)	(0.089)	(0.078)	(0.032)	(0.029)	(0.039)	(0.036)	(0.041)	(0.034)
Monitoring x Female Teacher		-0.063	-0.160**	-0.007	-0.052		-0.066		0.036	-0.020	-0.028
	(0.000)	(0.109)	(0.077)	(0.098)	(0.091)	(0.000)	(0.055)	(0.000)	(0.060)	(0.054)	(0.049)
Observations	439	439	439	438	419	418	418	413	416	2,721	3,801
R2	0.020	0.017	0.019	0.016	0.015	0.013	0.024	0.021	0.017	0.003	0.001
Adjusted R2	0.013	0.008	0.010	0.007	0.006	0.003	0.012	0.012	0.007	0.002	0.001
Residual Std. Error	0.491 (df = 435)	0.493 (df = 434)	0.492 (df = 434)	0.493 (df = 433)	0.492 (df = 414)	0.492 (df = 413)	0.491 (df = 412)	0.491 (df = 408)	0.492 (df = 411)	0.495 (df = 2717)	0.494 (df = 3797)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

Morality book

	Dependent variable:
	Morality book
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.048	-0.071	-0.035	-0.023	-0.066	-0.043	-0.065*	-0.060*	-0.0003	-0.059	-0.051
	(0.034)	(0.053)	(0.028)	(0.036)	(0.047)	(0.036)	(0.037)	(0.033)	(0.021)	(0.044)	(0.041)
Female Teacher	0.094***	-0.071	0.015	0.048	-0.053	0.062*	-0.069**	0.070***	0.074	0.021	0.005
	(0.025)	(0.043)	(0.046)	(0.084)	(0.051)	(0.032)	(0.030)	(0.026)	(0.058)	(0.033)	(0.027)
Monitoring x Female Teacher		0.111	0.015	-0.057	0.093*		0.169***		-0.150**	-0.001	0.013
	(0.000)	(0.072)	(0.060)	(0.090)	(0.055)	(0.000)	(0.044)	(0.000)	(0.074)	(0.042)	(0.038)
Observations	439	439	439	438	419	418	418	413	416	2,721	3,801
R2	0.017	0.018	0.015	0.016	0.014	0.011	0.016	0.019	0.022	0.006	0.003
Adjusted R2	0.010	0.009	0.006	0.007	0.004	0.002	0.004	0.009	0.012	0.005	0.003
Residual Std. Error	0.381 (df = 435)	0.381 (df = 434)	0.382 (df = 434)	0.382 (df = 433)	0.387 (df = 414)	0.390 (df = 413)	0.389 (df = 412)	0.384 (df = 408)	0.384 (df = 411)	0.380 (df = 2717)	0.385 (df = 3797)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

Judicial independence book

	Dependent variable:
	Judicial independence book
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.006	0.005	-0.033	-0.005	-0.020	-0.004	-0.007	0.016	0.023	0.030	0.019
	(0.034)	(0.048)	(0.037)	(0.035)	(0.043)	(0.037)	(0.044)	(0.039)	(0.037)	(0.055)	(0.051)
Female Teacher	0.084**	0.085**	-0.064*	0.009	0.037	-0.267***	0.131***	0.076***	-0.020	0.023	0.016
	(0.034)	(0.040)	(0.036)	(0.063)	(0.051)	(0.023)	(0.030)	(0.028)	(0.041)	(0.041)	(0.033)
Monitoring x Female Teacher		-0.059	0.123**	0.075	0.072		-0.023		0.056	0.023	0.026
	(0.000)	(0.064)	(0.050)	(0.070)	(0.062)	(0.000)	(0.060)	(0.000)	(0.052)	(0.053)	(0.048)
Observations	439	439	439	438	419	418	418	413	416	2,721	3,801
R2	0.004	0.005	0.006	0.006	0.007	0.015	0.011	0.008	0.004	0.003	0.001
Adjusted R2	-0.003	-0.004	-0.003	-0.003	-0.003	0.005	-0.001	-0.002	-0.005	0.002	0.001
Residual Std. Error	0.470 (df = 435)	0.469 (df = 434)	0.469 (df = 434)	0.469 (df = 433)	0.471 (df = 414)	0.470 (df = 413)	0.470 (df = 412)	0.471 (df = 408)	0.472 (df = 411)	0.470 (df = 2717)	0.469 (df = 3797)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.

Effect on exit survey answers

outcome_cols = c('work_satisfied_likert' = 'Satisfied with work',
                 'work_important_likert' = 'Work is important',
                 'motivated_likert' = 'Motivated to become better judge/prosecutor',
                 'trust_likert' = 'Trust in judicial system',
                 'learning_satisfied_likert' = 'Satisfaction with learning',
                 'improve_likert' = 'Training is key for improvement of judicial system',
                 'professor_likert' = 'Professors provide motivation',
                 'growth_1_likert' = 'Growth 1: I truly think people can change',
                 'growth_2_likert' = 'Growth 2: I can always change how smart I am',
                 "avg_growth_likert" = 'Average growth',
                 'challenge_1_likert' = 'Challenge 1: I enjoy projects that take years to be completed',
                 'challenge_2_likert' = 'Challenge 2: What I do everyday is connected to my deepest personal values',
                 'challenge_3_likert' = 'Challenge 3: There is at least one topic or activity that I never get bored thinking about',
                 'challenge_4_likert' = 'Challenge 4: I am hardworking',
                 "avg_challenge_likert" = 'Average challenge',
                 "empowerment_1_likert" = 'Empowerment 1: I question my own opinions, positions and points of view because they could be wrong',
                 "empowerment_2_likert" = "Empowerment 2: I recognize the value of opinions that are different than mine",
                 'avg_empowerment_likert' = 'Average empowerment',
                 "curiosity_1_likert" = 'Curiosity 1: I was so absorbed in learning that I lost track of time',
                 "curiosity_2_likert" = 'Curiosity 2: I explored a completely new idea or topic just for fun',
                 'avg_curiosity_likert' = 'Average curiosity',
                 "goal_1_likert" = 'Goal setting 1: I reflect on my life goals and the type of person I want to be',
                 "goal_2_likert" = 'Goal setting 2: When planning  my day, I think about how my activities are related to what I want to achieve in life',
                 'avg_goal_likert' = 'Average goal setting',
                 "gratitude_1_likert" = 'Gratitude 1: I say "Thank you"',
                 "gratitude_2_likert" = 'Gratitude 2: I did something nice to show my appreciation',
                 "avg_gratitude_likert" = 'Average gratitude')

note <-  "Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale."

for (outcome in names(outcome_cols)) {
  
  formula <- as.formula(paste0(outcome, ' ~ monitoreo * teacher_female | idcurso | 0 | aula'))
  
  out_list <- list()
  
  for (r in 1:9) {
    model <- felm(formula, data = data %>% filter(round == r))
    
    out_list[[sprintf('Round %i', r)]] <- model
  }
  
  formula <- as.formula(paste0(outcome, ' ~ monitoreo * teacher_female | 0 | 0 | dni'))
  
  out_list[['Pure Treatment']] <- felm(formula, data = pure_control)
  out_list[['Perc. courses monitored']] <- felm(formula, data = monit_perc)
  

  out_name <- outcome_cols[[outcome]]
  cat('<h4></h4>')
  cat(sprintf('<h3>%s</h3>', out_name))
  
  stargazer(out_list,
            covariate.labels = c('Monitoring', 'Female Teacher', 'Monitoring x Female Teacher'),
            dep.var.labels = out_name,
            column.labels = names(out_list),
            notes = note,
            omit = 'Constant',
            type = 'html')
  
  feols(fml = as.formula(sprintf('%s ~ i(round, monitoreo) | idcurso', outcome)),
                       data = data, cluster = ~aula) %>%
    iplot(dict = outcome_cols, xlab = 'Round') 
  
}

Satisfied with work

	Dependent variable:
	Satisfied with work
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.015	-0.007	0.001	-0.035	0.030	0.029	0.061	0.046	0.009	0.021	0.021
	(0.036)	(0.048)	(0.047)	(0.032)	(0.041)	(0.055)	(0.060)	(0.063)	(0.035)	(0.072)	(0.068)
Female Teacher	-0.019	0.095**	-0.036	-0.075	0.015	0.082*	0.210***	0.031	-0.049	0.009	-0.001
	(0.026)	(0.039)	(0.051)	(0.059)	(0.069)	(0.045)	(0.042)	(0.053)	(0.035)	(0.047)	(0.039)
Monitoring x Female Teacher		-0.038	-0.034	0.176**	-0.033		-0.215***		0.075	0.014	0.027
	(0.000)	(0.086)	(0.064)	(0.077)	(0.079)	(0.000)	(0.068)	(0.000)	(0.072)	(0.065)	(0.061)
Observations	459	459	459	458	439	438	438	433	436	2,864	3,978
R2	0.002	0.004	0.003	0.005	0.001	0.002	0.006	0.003	0.001	0.001	0.0005
Adjusted R2	-0.005	-0.005	-0.006	-0.004	-0.008	-0.008	-0.006	-0.006	-0.008	-0.001	-0.0003
Residual Std. Error	0.658 (df = 455)	0.658 (df = 454)	0.658 (df = 454)	0.658 (df = 453)	0.665 (df = 434)	0.666 (df = 433)	0.665 (df = 432)	0.660 (df = 428)	0.667 (df = 431)	0.631 (df = 2860)	0.659 (df = 3974)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Work is important

	Dependent variable:
	Work is important
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.014	-0.026	-0.006	-0.046	0.030	0.024	0.059	0.044	0.024	0.023	0.019
	(0.047)	(0.061)	(0.054)	(0.041)	(0.060)	(0.061)	(0.063)	(0.066)	(0.033)	(0.074)	(0.070)
Female Teacher	0.046	0.098*	-0.050	-0.106	0.023	0.0005	0.216***	0.078	-0.119***	-0.007	-0.013
	(0.041)	(0.051)	(0.060)	(0.069)	(0.076)	(0.049)	(0.046)	(0.058)	(0.035)	(0.048)	(0.040)
Monitoring x Female Teacher		0.015	0.012	0.242**	-0.002		-0.174**		0.129*	0.049	0.066
	(0.000)	(0.102)	(0.083)	(0.098)	(0.090)	(0.000)	(0.069)	(0.000)	(0.075)	(0.066)	(0.062)
Observations	459	459	459	458	439	438	438	433	436	2,864	3,978
R2	0.002	0.006	0.002	0.008	0.001	0.001	0.006	0.003	0.004	0.001	0.001
Adjusted R2	-0.005	-0.002	-0.006	-0.001	-0.008	-0.009	-0.005	-0.006	-0.005	0.0001	0.0003
Residual Std. Error	0.669 (df = 455)	0.669 (df = 454)	0.670 (df = 454)	0.669 (df = 453)	0.678 (df = 434)	0.679 (df = 433)	0.678 (df = 432)	0.674 (df = 428)	0.678 (df = 431)	0.648 (df = 2860)	0.671 (df = 3974)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Motivated to become better judge/prosecutor

	Dependent variable:
	Motivated to become better judge/prosecutor
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.006	-0.016	-0.038	-0.054	0.022	0.007	0.049	0.030	0.015	0.008	0.006
	(0.045)	(0.055)	(0.054)	(0.037)	(0.051)	(0.059)	(0.062)	(0.064)	(0.038)	(0.074)	(0.071)
Female Teacher	0.027	0.085**	-0.077	-0.114	0.007	-0.079	0.200***	0.069	-0.089**	-0.016	-0.024
	(0.049)	(0.042)	(0.058)	(0.071)	(0.080)	(0.048)	(0.047)	(0.054)	(0.040)	(0.048)	(0.040)
Monitoring x Female Teacher		-0.010	0.077	0.225**	-0.004		-0.173**		0.075	0.051	0.064
	(0.000)	(0.092)	(0.068)	(0.098)	(0.093)	(0.000)	(0.068)	(0.000)	(0.073)	(0.067)	(0.062)
Observations	459	459	459	458	439	438	438	433	436	2,864	3,978
R2	0.001	0.003	0.002	0.006	0.0005	0.001	0.004	0.003	0.002	0.001	0.001
Adjusted R2	-0.006	-0.005	-0.006	-0.003	-0.009	-0.009	-0.007	-0.007	-0.008	-0.001	-0.0002
Residual Std. Error	0.676 (df = 455)	0.676 (df = 454)	0.676 (df = 454)	0.675 (df = 453)	0.684 (df = 434)	0.685 (df = 433)	0.684 (df = 432)	0.678 (df = 428)	0.686 (df = 431)	0.651 (df = 2860)	0.677 (df = 3974)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Trust in judicial system

	Dependent variable:
	Trust in judicial system
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.017	-0.036	-0.044	-0.064**	-0.016	-0.009	0.030	0.004	-0.043**	-0.035	-0.021
	(0.040)	(0.050)	(0.039)	(0.027)	(0.053)	(0.062)	(0.065)	(0.071)	(0.021)	(0.068)	(0.065)
Female Teacher	0.038	0.100***	-0.057	-0.081	0.012	-0.027	0.255***	0.054	-0.086	-0.018	-0.016
	(0.027)	(0.036)	(0.051)	(0.083)	(0.064)	(0.051)	(0.047)	(0.059)	(0.090)	(0.047)	(0.039)
Monitoring x Female Teacher		-0.027	0.050	0.210**	-0.006		-0.226***		0.169*	0.068	0.070
	(0.000)	(0.090)	(0.060)	(0.095)	(0.073)	(0.000)	(0.071)	(0.000)	(0.102)	(0.065)	(0.060)
Observations	458	458	458	457	438	437	437	432	435	2,864	3,969
R2	0.003	0.007	0.004	0.008	0.003	0.003	0.010	0.005	0.004	0.001	0.001
Adjusted R2	-0.003	-0.002	-0.005	-0.0004	-0.006	-0.006	-0.002	-0.004	-0.005	-0.0001	-0.0002
Residual Std. Error	0.637 (df = 454)	0.636 (df = 453)	0.637 (df = 453)	0.636 (df = 452)	0.643 (df = 433)	0.643 (df = 432)	0.641 (df = 431)	0.639 (df = 427)	0.643 (df = 430)	0.609 (df = 2860)	0.637 (df = 3965)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Satisfaction with learning

	Dependent variable:
	Satisfaction with learning
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.044	0.033	0.041	0.005	0.075	0.064	0.095	0.069	0.069**	0.060	0.075
	(0.039)	(0.057)	(0.050)	(0.038)	(0.047)	(0.052)	(0.059)	(0.057)	(0.033)	(0.074)	(0.070)
Female Teacher	0.027	0.078*	-0.018	-0.057	0.047	0.043	0.165***	0.082*	-0.087***	0.002	0.0004
	(0.034)	(0.040)	(0.050)	(0.055)	(0.058)	(0.042)	(0.043)	(0.048)	(0.024)	(0.050)	(0.042)
Monitoring x Female Teacher		-0.039	-0.004	0.171**	-0.032		-0.150**		0.068	0.020	0.027
	(0.000)	(0.085)	(0.066)	(0.067)	(0.073)	(0.000)	(0.072)	(0.000)	(0.056)	(0.069)	(0.064)
Observations	454	454	454	453	435	434	434	429	432	2,847	3,939
R2	0.001	0.002	0.002	0.005	0.003	0.003	0.007	0.007	0.005	0.003	0.003
Adjusted R2	-0.005	-0.007	-0.007	-0.004	-0.006	-0.006	-0.005	-0.003	-0.005	0.002	0.002
Residual Std. Error	0.678 (df = 450)	0.678 (df = 449)	0.679 (df = 449)	0.678 (df = 448)	0.678 (df = 430)	0.678 (df = 429)	0.678 (df = 428)	0.672 (df = 424)	0.682 (df = 427)	0.650 (df = 2843)	0.675 (df = 3935)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Training is key for improvement of judicial system

	Dependent variable:
	Training is key for improvement of judicial system
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.008	-0.022	-0.037	-0.047	0.032	0.020	0.062	0.042	0.025	0.012	0.015
	(0.044)	(0.056)	(0.057)	(0.041)	(0.050)	(0.057)	(0.060)	(0.061)	(0.037)	(0.074)	(0.071)
Female Teacher	0.034	0.083	-0.081	-0.127*	0.038	-0.044	0.200***	0.035	-0.093**	-0.012	-0.017
	(0.034)	(0.055)	(0.064)	(0.070)	(0.090)	(0.049)	(0.049)	(0.050)	(0.046)	(0.047)	(0.039)
Monitoring x Female Teacher		0.011	0.083	0.226***	-0.019		-0.189***		0.064	0.042	0.054
	(0.000)	(0.090)	(0.073)	(0.085)	(0.096)	(0.000)	(0.070)	(0.000)	(0.082)	(0.066)	(0.062)
Observations	460	460	460	459	439	438	438	433	436	2,864	3,982
R2	0.001	0.004	0.002	0.006	0.001	0.0004	0.005	0.002	0.002	0.001	0.001
Adjusted R2	-0.006	-0.005	-0.006	-0.003	-0.008	-0.009	-0.007	-0.007	-0.007	-0.001	-0.0002
Residual Std. Error	0.675 (df = 456)	0.675 (df = 455)	0.675 (df = 455)	0.674 (df = 454)	0.684 (df = 434)	0.684 (df = 433)	0.684 (df = 432)	0.678 (df = 428)	0.685 (df = 431)	0.650 (df = 2860)	0.677 (df = 3978)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Professors provide motivation

	Dependent variable:
	Professors provide motivation
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.059	0.029	0.077*	0.016	0.091	0.081	0.134**	0.105	0.052	0.095	0.095
	(0.043)	(0.057)	(0.045)	(0.043)	(0.061)	(0.057)	(0.064)	(0.065)	(0.051)	(0.073)	(0.069)
Female Teacher	-0.088***	0.116***	0.021	-0.044	0.041	0.063	0.218***	0.017	-0.129***	0.025	0.025
	(0.032)	(0.042)	(0.059)	(0.078)	(0.065)	(0.049)	(0.046)	(0.061)	(0.030)	(0.049)	(0.041)
Monitoring x Female Teacher		-0.066	-0.102	0.148	-0.058		-0.275***		0.166**	-0.023	-0.017
	(0.000)	(0.092)	(0.063)	(0.091)	(0.073)	(0.000)	(0.086)	(0.000)	(0.077)	(0.069)	(0.064)
Observations	459	459	459	458	439	438	438	433	436	2,864	3,978
R2	0.002	0.004	0.003	0.005	0.004	0.004	0.011	0.007	0.009	0.005	0.004
Adjusted R2	-0.004	-0.005	-0.006	-0.004	-0.006	-0.006	-0.001	-0.002	-0.0005	0.004	0.003
Residual Std. Error	0.667 (df = 455)	0.667 (df = 454)	0.667 (df = 454)	0.667 (df = 453)	0.668 (df = 434)	0.668 (df = 433)	0.667 (df = 432)	0.664 (df = 428)	0.668 (df = 431)	0.646 (df = 2860)	0.664 (df = 3974)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Growth 1: I truly think people can change

	Dependent variable:
	Growth 1: I truly think people can change
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.069*	-0.076	-0.052	-0.086*	-0.074	-0.052	-0.042	-0.050	-0.005	-0.078	-0.061
	(0.039)	(0.062)	(0.046)	(0.044)	(0.046)	(0.042)	(0.044)	(0.044)	(0.032)	(0.051)	(0.049)
Female Teacher	0.078**	-0.005	0.033	-0.029	-0.074**	0.058*	0.151***	0.054	0.013	0.018	0.026
	(0.030)	(0.030)	(0.034)	(0.048)	(0.035)	(0.034)	(0.027)	(0.045)	(0.065)	(0.036)	(0.030)
Monitoring x Female Teacher		0.010	0.028	0.107	0.060		-0.080		-0.010	0.007	-0.002
	(0.000)	(0.081)	(0.048)	(0.069)	(0.077)	(0.000)	(0.049)	(0.000)	(0.079)	(0.052)	(0.049)
Observations	453	453	453	452	433	432	432	427	430	2,847	3,926
R2	0.007	0.008	0.008	0.009	0.012	0.010	0.013	0.008	0.011	0.007	0.004
Adjusted R2	0.001	-0.001	-0.001	0.0004	0.002	0.0003	0.002	-0.001	0.002	0.006	0.003
Residual Std. Error	0.461 (df = 449)	0.462 (df = 448)	0.462 (df = 448)	0.462 (df = 447)	0.460 (df = 428)	0.461 (df = 427)	0.461 (df = 426)	0.457 (df = 422)	0.467 (df = 425)	0.452 (df = 2843)	0.462 (df = 3922)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Growth 2: I can always change how smart I am

	Dependent variable:
	Growth 2: I can always change how smart I am
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.085*	-0.052	-0.116**	-0.060	-0.099*	-0.048	-0.049	-0.059	0.060	-0.044	-0.049
	(0.047)	(0.081)	(0.059)	(0.054)	(0.059)	(0.054)	(0.055)	(0.057)	(0.040)	(0.059)	(0.056)
Female Teacher	0.150***	0.065	-0.032	0.015	-0.086	0.085**	0.195***	0.104*	0.136*	0.044	0.056*
	(0.036)	(0.053)	(0.058)	(0.082)	(0.065)	(0.043)	(0.033)	(0.058)	(0.075)	(0.039)	(0.032)
Monitoring x Female Teacher		-0.070	0.185***	-0.058	0.048		-0.103*		-0.211**	-0.038	-0.056
	(0.000)	(0.097)	(0.068)	(0.092)	(0.080)	(0.000)	(0.062)	(0.000)	(0.093)	(0.055)	(0.052)
Observations	445	445	445	444	425	425	424	419	422	2,793	3,855
R2	0.016	0.014	0.019	0.014	0.016	0.015	0.022	0.016	0.020	0.003	0.003
Adjusted R2	0.009	0.005	0.010	0.005	0.006	0.005	0.010	0.006	0.011	0.002	0.003
Residual Std. Error	0.529 (df = 441)	0.530 (df = 440)	0.529 (df = 440)	0.530 (df = 439)	0.525 (df = 420)	0.527 (df = 420)	0.526 (df = 418)	0.523 (df = 414)	0.530 (df = 417)	0.514 (df = 2789)	0.529 (df = 3851)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Average growth

	Dependent variable:
	Average growth
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.077*	-0.066	-0.086*	-0.074	-0.087*	-0.053	-0.046	-0.057	0.022	-0.063	-0.056
	(0.040)	(0.066)	(0.049)	(0.046)	(0.048)	(0.045)	(0.046)	(0.049)	(0.034)	(0.050)	(0.047)
Female Teacher	0.092***	0.029	-0.001	-0.008	-0.080*	0.071*	0.172***	0.072	0.074	0.030	0.042
	(0.029)	(0.039)	(0.044)	(0.064)	(0.048)	(0.036)	(0.028)	(0.050)	(0.070)	(0.034)	(0.028)
Monitoring x Female Teacher		-0.031	0.099*	0.020	0.046		-0.106**		-0.103	-0.020	-0.034
	(0.000)	(0.080)	(0.054)	(0.075)	(0.068)	(0.000)	(0.051)	(0.000)	(0.083)	(0.050)	(0.047)
Observations	445	445	445	444	425	425	424	419	422	2,793	3,855
R2	0.014	0.013	0.015	0.012	0.018	0.015	0.021	0.015	0.018	0.006	0.005
Adjusted R2	0.007	0.004	0.006	0.003	0.008	0.005	0.009	0.005	0.008	0.005	0.004
Residual Std. Error	0.450 (df = 441)	0.451 (df = 440)	0.451 (df = 440)	0.452 (df = 439)	0.450 (df = 420)	0.451 (df = 420)	0.450 (df = 418)	0.445 (df = 414)	0.456 (df = 417)	0.439 (df = 2789)	0.452 (df = 3851)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Challenge 1: I enjoy projects that take years to be completed

	Dependent variable:
	Challenge 1: I enjoy projects that take years to be completed
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.047	0.017	-0.121	-0.018	-0.063	-0.041	-0.077	-0.093	0.055	-0.065	-0.072
	(0.069)	(0.094)	(0.099)	(0.062)	(0.084)	(0.068)	(0.068)	(0.066)	(0.053)	(0.069)	(0.065)
Female Teacher	0.084	-0.053	-0.099	-0.160***	-0.097	-0.081	-0.117**	0.175***	0.027	-0.103**	-0.091**
	(0.057)	(0.074)	(0.082)	(0.058)	(0.086)	(0.051)	(0.051)	(0.061)	(0.207)	(0.048)	(0.040)
Monitoring x Female Teacher		-0.072	0.259**	-0.101	0.043		0.287***		-0.211	0.069	0.074
	(0.000)	(0.154)	(0.109)	(0.137)	(0.140)	(0.000)	(0.075)	(0.000)	(0.217)	(0.067)	(0.063)
Observations	451	451	451	450	431	430	430	426	429	2,838	3,910
R2	0.002	0.006	0.009	0.022	0.006	0.004	0.009	0.008	0.012	0.004	0.003
Adjusted R2	-0.004	-0.003	0.0001	0.013	-0.004	-0.005	-0.003	-0.002	0.002	0.003	0.003
Residual Std. Error	0.636 (df = 447)	0.637 (df = 446)	0.636 (df = 446)	0.632 (df = 445)	0.637 (df = 426)	0.636 (df = 425)	0.634 (df = 424)	0.634 (df = 421)	0.631 (df = 424)	0.612 (df = 2834)	0.633 (df = 3906)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Challenge 2: What I do everyday is connected to my deepest personal values

	Dependent variable:
	Challenge 2: What I do everyday is connected to my deepest personal values
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.050	-0.077	-0.062	-0.071	-0.075	-0.029	-0.030	-0.028	0.068**	-0.037	-0.036
	(0.038)	(0.057)	(0.045)	(0.046)	(0.050)	(0.047)	(0.052)	(0.045)	(0.030)	(0.049)	(0.047)
Female Teacher	0.077*	0.034	-0.022	-0.065	-0.051	0.017	0.149***	0.038	0.079	-0.001	0.016
	(0.041)	(0.062)	(0.041)	(0.042)	(0.044)	(0.036)	(0.032)	(0.045)	(0.063)	(0.032)	(0.027)
Monitoring x Female Teacher		0.030	0.047	0.107*	0.074		-0.120**		-0.118	0.006	0.0002
	(0.000)	(0.087)	(0.061)	(0.055)	(0.062)	(0.000)	(0.055)	(0.000)	(0.082)	(0.051)	(0.048)
Observations	450	450	450	449	430	429	429	425	427	2,838	3,900
R2	0.004	0.007	0.003	0.006	0.005	0.006	0.011	0.005	0.007	0.002	0.001
Adjusted R2	-0.003	-0.002	-0.005	-0.003	-0.004	-0.003	-0.001	-0.005	-0.003	0.001	0.001
Residual Std. Error	0.454 (df = 446)	0.453 (df = 445)	0.454 (df = 445)	0.454 (df = 444)	0.453 (df = 425)	0.454 (df = 424)	0.453 (df = 423)	0.448 (df = 420)	0.462 (df = 422)	0.446 (df = 2834)	0.454 (df = 3896)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Challenge 3: There is at least one topic or activity that I never get bored thinking about

	Dependent variable:
	Challenge 3: There is at least one topic or activity that I never get bored thinking about
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.074*	-0.099*	-0.110**	-0.073*	-0.106**	-0.125***	-0.121***	-0.138***	-0.024	-0.128**	-0.109*
	(0.039)	(0.056)	(0.047)	(0.043)	(0.053)	(0.040)	(0.037)	(0.040)	(0.043)	(0.060)	(0.056)
Female Teacher	0.008	0.011	-0.070	0.019	-0.062	-0.074*	0.175***	0.074**	0.174**	-0.016	-0.007
	(0.033)	(0.046)	(0.057)	(0.074)	(0.063)	(0.042)	(0.035)	(0.030)	(0.081)	(0.035)	(0.029)
Monitoring x Female Teacher		0.041	0.043	-0.006	0.067		-0.127***		-0.168*	0.029	0.023
	(0.000)	(0.071)	(0.063)	(0.078)	(0.065)	(0.000)	(0.045)	(0.000)	(0.086)	(0.057)	(0.053)
Observations	452	452	452	451	432	431	431	426	429	2,838	3,917
R2	0.011	0.013	0.014	0.011	0.013	0.031	0.036	0.035	0.012	0.014	0.008
Adjusted R2	0.004	0.004	0.005	0.002	0.003	0.021	0.025	0.026	0.003	0.012	0.007
Residual Std. Error	0.524 (df = 448)	0.524 (df = 447)	0.524 (df = 447)	0.525 (df = 446)	0.525 (df = 427)	0.521 (df = 426)	0.520 (df = 425)	0.519 (df = 421)	0.530 (df = 424)	0.525 (df = 2834)	0.525 (df = 3913)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Challenge 4: I am hardworking

	Dependent variable:
	Challenge 4: I am hardworking
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.039	-0.103*	-0.085	-0.074	-0.052	-0.037	-0.034	-0.058	0.021	-0.051	-0.052
	(0.048)	(0.060)	(0.053)	(0.049)	(0.064)	(0.052)	(0.057)	(0.058)	(0.050)	(0.049)	(0.047)
Female Teacher	0.053	-0.047	-0.085*	0.010	-0.041	-0.093***	0.122***	0.062	0.111*	-0.004	-0.008
	(0.058)	(0.035)	(0.049)	(0.068)	(0.052)	(0.034)	(0.032)	(0.059)	(0.067)	(0.031)	(0.026)
Monitoring x Female Teacher		0.172*	0.101	0.158*	0.055		-0.068		-0.100	0.049	0.064
	(0.000)	(0.102)	(0.071)	(0.091)	(0.080)	(0.000)	(0.060)	(0.000)	(0.084)	(0.050)	(0.047)
Observations	449	449	449	448	429	428	428	423	426	2,838	3,890
R2	0.002	0.010	0.006	0.014	0.003	0.005	0.008	0.007	0.005	0.003	0.003
Adjusted R2	-0.005	0.001	-0.003	0.006	-0.007	-0.004	-0.004	-0.003	-0.004	0.002	0.002
Residual Std. Error	0.447 (df = 445)	0.446 (df = 444)	0.447 (df = 444)	0.445 (df = 443)	0.454 (df = 424)	0.454 (df = 423)	0.454 (df = 422)	0.448 (df = 418)	0.454 (df = 421)	0.438 (df = 2834)	0.450 (df = 3886)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Average challenge

	Dependent variable:
	Average challenge
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.055	-0.069	-0.099**	-0.068	-0.083*	-0.062	-0.066	-0.080**	0.020	-0.070	-0.072*
	(0.036)	(0.054)	(0.048)	(0.043)	(0.048)	(0.040)	(0.041)	(0.039)	(0.033)	(0.043)	(0.041)
Female Teacher	0.037	-0.015	-0.070	-0.050	-0.064	-0.096***	0.082***	0.095***	0.096	-0.031	-0.023
	(0.035)	(0.042)	(0.046)	(0.049)	(0.048)	(0.032)	(0.030)	(0.036)	(0.101)	(0.027)	(0.023)
Monitoring x Female Teacher		0.049	0.106*	0.056	0.073		-0.009		-0.141	0.042	0.044
	(0.000)	(0.067)	(0.055)	(0.058)	(0.057)	(0.000)	(0.044)	(0.000)	(0.106)	(0.045)	(0.042)
Observations	443	443	443	442	423	422	422	419	421	2,811	3,839
R2	0.005	0.006	0.010	0.007	0.009	0.010	0.012	0.014	0.007	0.007	0.006
Adjusted R2	-0.001	-0.003	0.001	-0.002	-0.001	0.0002	-0.0001	0.004	-0.003	0.006	0.005
Residual Std. Error	0.396 (df = 439)	0.396 (df = 438)	0.395 (df = 438)	0.396 (df = 437)	0.398 (df = 418)	0.398 (df = 417)	0.398 (df = 416)	0.393 (df = 414)	0.400 (df = 416)	0.383 (df = 2807)	0.396 (df = 3835)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Empowerment 1: I question my own opinions, positions and points of view because they could be wrong

	Dependent variable:
	Empowerment 1: I question my own opinions, positions and points of view because they could be wrong
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.028	0.033	0.034	0.024	0.015	0.055	0.024	0.021	0.130***	0.048	0.045
	(0.043)	(0.056)	(0.052)	(0.047)	(0.052)	(0.046)	(0.045)	(0.042)	(0.040)	(0.066)	(0.063)
Female Teacher	0.134***	-0.008	0.008	-0.087	-0.106*	-0.028	-0.021	-0.006	0.085**	-0.070	-0.055
	(0.045)	(0.092)	(0.051)	(0.074)	(0.059)	(0.034)	(0.034)	(0.033)	(0.037)	(0.053)	(0.043)
Monitoring x Female Teacher		-0.021	0.044	0.052	0.098		0.034		-0.168**	0.079	0.060
	(0.000)	(0.108)	(0.079)	(0.085)	(0.071)	(0.000)	(0.080)	(0.000)	(0.069)	(0.069)	(0.063)
Observations	448	448	448	447	428	427	427	422	425	2,820	3,881
R2	0.009	0.006	0.007	0.009	0.011	0.025	0.022	0.018	0.011	0.004	0.002
Adjusted R2	0.002	-0.002	-0.002	0.00005	0.002	0.016	0.011	0.008	0.002	0.003	0.002
Residual Std. Error	0.607 (df = 444)	0.609 (df = 443)	0.608 (df = 443)	0.609 (df = 442)	0.608 (df = 423)	0.601 (df = 422)	0.605 (df = 421)	0.608 (df = 417)	0.606 (df = 420)	0.595 (df = 2816)	0.607 (df = 3877)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Empowerment 2: I recognize the value of opinions that are different than mine

	Dependent variable:
	Empowerment 2: I recognize the value of opinions that are different than mine
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.028	-0.063	-0.066	-0.051	-0.042	-0.023	-0.014	-0.016	0.082**	-0.004	-0.020
	(0.045)	(0.064)	(0.054)	(0.049)	(0.055)	(0.050)	(0.054)	(0.050)	(0.041)	(0.049)	(0.046)
Female Teacher	0.006	-0.025	-0.067	-0.092	-0.089	0.010	0.068*	-0.050	0.124*	-0.025	-0.021
	(0.046)	(0.077)	(0.057)	(0.081)	(0.063)	(0.038)	(0.038)	(0.046)	(0.065)	(0.032)	(0.026)
Monitoring x Female Teacher		0.070	0.083	0.093	0.065		-0.080		-0.150*	0.013	0.019
	(0.000)	(0.101)	(0.067)	(0.093)	(0.078)	(0.000)	(0.063)	(0.000)	(0.079)	(0.050)	(0.046)
Observations	450	450	450	449	429	428	428	423	426	2,820	3,894
R2	0.002	0.004	0.005	0.006	0.005	0.010	0.011	0.012	0.009	0.0003	0.0004
Adjusted R2	-0.005	-0.005	-0.004	-0.003	-0.005	0.001	-0.001	0.003	-0.001	-0.001	-0.0003
Residual Std. Error	0.440 (df = 446)	0.440 (df = 445)	0.439 (df = 445)	0.439 (df = 444)	0.446 (df = 424)	0.445 (df = 423)	0.446 (df = 422)	0.439 (df = 418)	0.446 (df = 421)	0.435 (df = 2816)	0.443 (df = 3890)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Average empowerment

	Dependent variable:
	Average empowerment
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.001	-0.013	-0.016	-0.008	-0.009	0.019	0.005	0.006	0.113***	0.026	0.015
	(0.040)	(0.059)	(0.049)	(0.044)	(0.052)	(0.045)	(0.047)	(0.045)	(0.034)	(0.051)	(0.047)
Female Teacher	0.116***	-0.015	-0.029	-0.087	-0.095	-0.007	0.026	-0.029	0.107**	-0.048	-0.038
	(0.033)	(0.084)	(0.051)	(0.077)	(0.058)	(0.034)	(0.035)	(0.039)	(0.049)	(0.037)	(0.029)
Monitoring x Female Teacher		0.027	0.085	0.065	0.077		-0.004		-0.167**	0.052	0.046
	(0.000)	(0.103)	(0.069)	(0.085)	(0.071)	(0.000)	(0.076)	(0.000)	(0.068)	(0.053)	(0.048)
Observations	446	446	446	445	426	425	425	420	423	2,802	3,863
R2	0.008	0.006	0.007	0.009	0.010	0.022	0.022	0.021	0.014	0.002	0.001
Adjusted R2	0.001	-0.003	-0.002	0.0004	0.0003	0.013	0.010	0.011	0.004	0.001	0.0003
Residual Std. Error	0.450 (df = 442)	0.451 (df = 441)	0.451 (df = 441)	0.451 (df = 440)	0.454 (df = 421)	0.451 (df = 420)	0.452 (df = 419)	0.451 (df = 415)	0.452 (df = 418)	0.451 (df = 2798)	0.452 (df = 3859)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Curiosity 1: I was so absorbed in learning that I lost track of time

	Dependent variable:
	Curiosity 1: I was so absorbed in learning that I lost track of time
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.097*	-0.094	-0.127*	-0.085	0.010	-0.029	-0.039	-0.056	-0.053	-0.054	-0.071
	(0.059)	(0.064)	(0.072)	(0.065)	(0.062)	(0.061)	(0.062)	(0.059)	(0.064)	(0.067)	(0.063)
Female Teacher	0.155***	0.048	-0.083	-0.013	0.154**	-0.086*	0.268***	0.225***	-0.053	0.015	0.012
	(0.046)	(0.088)	(0.068)	(0.098)	(0.062)	(0.052)	(0.047)	(0.044)	(0.089)	(0.050)	(0.041)
Monitoring x Female Teacher		-0.070	0.056	-0.013	-0.253*		-0.091		0.125	0.010	0.022
	(0.000)	(0.116)	(0.131)	(0.147)	(0.131)	(0.000)	(0.066)	(0.000)	(0.119)	(0.065)	(0.059)
Observations	456	456	456	455	435	434	434	429	432	2,846	3,947
R2	0.008	0.011	0.009	0.007	0.013	0.005	0.018	0.011	0.003	0.002	0.003
Adjusted R2	0.002	0.002	0.0002	-0.002	0.004	-0.004	0.007	0.002	-0.007	0.001	0.002
Residual Std. Error	0.600 (df = 452)	0.598 (df = 451)	0.599 (df = 451)	0.599 (df = 450)	0.603 (df = 430)	0.604 (df = 429)	0.600 (df = 428)	0.601 (df = 424)	0.601 (df = 427)	0.592 (df = 2842)	0.599 (df = 3943)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Curiosity 2: I explored a completely new idea or topic just for fun

	Dependent variable:
	Curiosity 2: I explored a completely new idea or topic just for fun
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.069	-0.044	-0.037	-0.008	-0.037	0.010	-0.015	0.009	0.099*	-0.004	-0.023
	(0.046)	(0.050)	(0.072)	(0.057)	(0.055)	(0.051)	(0.048)	(0.048)	(0.053)	(0.061)	(0.057)
Female Teacher	0.279***	0.087*	0.043	0.112***	0.003	0.011	0.163***	0.030	0.099	0.033	0.029
	(0.056)	(0.046)	(0.067)	(0.042)	(0.066)	(0.035)	(0.030)	(0.038)	(0.072)	(0.041)	(0.035)
Monitoring x Female Teacher		-0.031	0.014	-0.169	-0.037		-0.007		-0.246**	-0.021	-0.002
	(0.000)	(0.071)	(0.151)	(0.109)	(0.117)	(0.000)	(0.074)	(0.000)	(0.110)	(0.057)	(0.052)
Observations	455	455	455	454	435	434	434	429	432	2,846	3,943
R2	0.011	0.005	0.005	0.008	0.003	0.004	0.014	0.004	0.013	0.0004	0.001
Adjusted R2	0.004	-0.004	-0.004	-0.001	-0.006	-0.005	0.002	-0.005	0.004	-0.001	0.00000
Residual Std. Error	0.557 (df = 451)	0.558 (df = 450)	0.558 (df = 450)	0.557 (df = 449)	0.562 (df = 430)	0.563 (df = 429)	0.559 (df = 428)	0.563 (df = 424)	0.556 (df = 427)	0.547 (df = 2842)	0.558 (df = 3939)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Average curiosity

	Dependent variable:
	Average curiosity
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.085*	-0.074	-0.085	-0.048	-0.015	-0.011	-0.030	-0.026	0.025	-0.031	-0.048
	(0.045)	(0.046)	(0.064)	(0.053)	(0.052)	(0.046)	(0.044)	(0.045)	(0.054)	(0.056)	(0.052)
Female Teacher	0.223***	0.068	-0.022	0.051	0.080	-0.036	0.217***	0.131***	0.024	0.025	0.022
	(0.045)	(0.060)	(0.058)	(0.060)	(0.052)	(0.036)	(0.030)	(0.036)	(0.079)	(0.041)	(0.033)
Monitoring x Female Teacher		-0.045	0.041	-0.089	-0.144		-0.045		-0.067	-0.004	0.011
	(0.000)	(0.079)	(0.136)	(0.120)	(0.112)	(0.000)	(0.053)	(0.000)	(0.104)	(0.054)	(0.048)
Observations	454	454	454	453	434	433	433	428	431	2,837	3,934
R2	0.012	0.009	0.007	0.008	0.008	0.006	0.022	0.009	0.002	0.001	0.002
Adjusted R2	0.006	0.0003	-0.002	-0.0004	-0.001	-0.004	0.010	-0.001	-0.007	0.0002	0.001
Residual Std. Error	0.495 (df = 450)	0.494 (df = 449)	0.495 (df = 449)	0.494 (df = 448)	0.502 (df = 429)	0.501 (df = 428)	0.498 (df = 427)	0.501 (df = 423)	0.499 (df = 426)	0.495 (df = 2833)	0.497 (df = 3930)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Goal setting 1: I reflect on my life goals and the type of person I want to be

	Dependent variable:
	Goal setting 1: I reflect on my life goals and the type of person I want to be
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.013	-0.016	0.074*	0.056	0.027	0.046	0.048	0.049	0.069*	0.049	0.051
	(0.033)	(0.041)	(0.042)	(0.043)	(0.043)	(0.036)	(0.034)	(0.038)	(0.041)	(0.047)	(0.044)
Female Teacher	0.201***	-0.006	0.093**	0.076*	-0.048	-0.029	0.058***	-0.030	-0.012	0.031	0.033
	(0.037)	(0.031)	(0.040)	(0.045)	(0.055)	(0.023)	(0.018)	(0.034)	(0.066)	(0.041)	(0.032)
Monitoring x Female Teacher		0.090	-0.041	-0.092	-0.014		0.0003		-0.028	-0.019	-0.023
	(0.000)	(0.062)	(0.098)	(0.059)	(0.066)	(0.000)	(0.093)	(0.000)	(0.084)	(0.053)	(0.047)
Observations	455	455	455	454	435	434	434	429	432	2,855	3,943
R2	0.010	0.005	0.009	0.005	0.005	0.009	0.011	0.007	0.008	0.003	0.003
Adjusted R2	0.003	-0.004	0.0002	-0.003	-0.004	-0.001	-0.001	-0.003	-0.001	0.002	0.002
Residual Std. Error	0.433 (df = 451)	0.435 (df = 450)	0.434 (df = 450)	0.435 (df = 449)	0.438 (df = 430)	0.438 (df = 429)	0.438 (df = 428)	0.435 (df = 424)	0.440 (df = 427)	0.426 (df = 2851)	0.435 (df = 3939)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Goal setting 2: When planning my day, I think about how my activities are related to what I want to achieve in life

	Dependent variable:
	Goal setting 2: When planning my day, I think about how my activities are related to what I want to achieve in life
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.014	-0.006	0.071	0.050	-0.009	0.039	0.051	0.051	0.089**	0.040	0.052
	(0.041)	(0.043)	(0.056)	(0.047)	(0.047)	(0.039)	(0.042)	(0.043)	(0.039)	(0.048)	(0.045)
Female Teacher	0.118*	0.022	0.072	0.100***	-0.048	-0.143***	0.136***	-0.012	0.069	0.050	0.054*
	(0.061)	(0.031)	(0.052)	(0.037)	(0.057)	(0.028)	(0.023)	(0.043)	(0.051)	(0.038)	(0.030)
Monitoring x Female Teacher		0.026	-0.106	-0.090	0.080		-0.095		-0.116*	-0.044	-0.054
	(0.000)	(0.069)	(0.107)	(0.066)	(0.076)	(0.000)	(0.059)	(0.000)	(0.060)	(0.050)	(0.045)
Observations	454	454	454	453	434	433	433	428	431	2,846	3,934
R2	0.008	0.006	0.009	0.010	0.006	0.011	0.013	0.008	0.016	0.002	0.003
Adjusted R2	0.001	-0.003	-0.0002	0.001	-0.003	0.002	0.001	-0.002	0.006	0.001	0.002
Residual Std. Error	0.435 (df = 450)	0.436 (df = 449)	0.435 (df = 449)	0.435 (df = 448)	0.439 (df = 429)	0.438 (df = 428)	0.438 (df = 427)	0.432 (df = 423)	0.440 (df = 426)	0.430 (df = 2842)	0.436 (df = 3930)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Average goal setting

	Dependent variable:
	Average goal setting
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	0.013	-0.011	0.072	0.053	0.008	0.042	0.049	0.049	0.079**	0.044	0.051
	(0.035)	(0.038)	(0.047)	(0.043)	(0.043)	(0.035)	(0.035)	(0.037)	(0.038)	(0.043)	(0.040)
Female Teacher	0.160***	0.008	0.082*	0.088**	-0.048	-0.086***	0.097***	-0.020	0.029	0.040	0.044
	(0.048)	(0.028)	(0.045)	(0.037)	(0.053)	(0.024)	(0.018)	(0.035)	(0.058)	(0.036)	(0.029)
Monitoring x Female Teacher		0.056	-0.073	-0.094*	0.034		-0.046		-0.073	-0.032	-0.039
	(0.000)	(0.060)	(0.101)	(0.048)	(0.066)	(0.000)	(0.072)	(0.000)	(0.070)	(0.048)	(0.043)
Observations	454	454	454	453	434	433	433	428	431	2,846	3,934
R2	0.010	0.006	0.009	0.009	0.005	0.010	0.013	0.008	0.014	0.003	0.003
Adjusted R2	0.003	-0.003	0.001	-0.0001	-0.004	0.001	0.002	-0.001	0.004	0.002	0.003
Residual Std. Error	0.393 (df = 450)	0.394 (df = 449)	0.393 (df = 449)	0.394 (df = 448)	0.399 (df = 429)	0.399 (df = 428)	0.399 (df = 427)	0.393 (df = 423)	0.399 (df = 426)	0.388 (df = 2842)	0.395 (df = 3930)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Gratitude 1: I say “Thank you”

	Dependent variable:
	Gratitude 1: I say “Thank you”
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.035	-0.077*	-0.011	-0.039	-0.067	0.027	0.039	0.033	0.080*	0.005	-0.002
	(0.038)	(0.045)	(0.042)	(0.045)	(0.050)	(0.036)	(0.041)	(0.037)	(0.048)	(0.044)	(0.041)
Female Teacher	0.100**	-0.018	0.018	0.018	-0.056	-0.073**	0.100***	0.029	0.143***	0.015	0.022
	(0.048)	(0.055)	(0.043)	(0.054)	(0.036)	(0.029)	(0.029)	(0.034)	(0.030)	(0.032)	(0.025)
Monitoring x Female Teacher		0.089	-0.051	0.045	0.120*		-0.065		-0.141**	0.005	0.006
	(0.000)	(0.076)	(0.095)	(0.072)	(0.061)	(0.000)	(0.051)	(0.000)	(0.065)	(0.045)	(0.041)
Observations	454	454	454	453	434	433	433	428	431	2,846	3,934
R2	0.003	0.006	0.002	0.004	0.007	0.010	0.007	0.008	0.015	0.0004	0.001
Adjusted R2	-0.003	-0.002	-0.007	-0.005	-0.003	0.0005	-0.004	-0.001	0.006	-0.001	-0.00003
Residual Std. Error	0.386 (df = 450)	0.386 (df = 449)	0.387 (df = 449)	0.387 (df = 448)	0.393 (df = 429)	0.392 (df = 428)	0.393 (df = 427)	0.384 (df = 423)	0.392 (df = 426)	0.385 (df = 2842)	0.389 (df = 3930)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Gratitude 2: I did something nice to show my appreciation

	Dependent variable:
	Gratitude 2: I did something nice to show my appreciation
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.065	-0.083	-0.040	-0.061	-0.089	-0.006	-0.021	-0.008	0.052	-0.007	-0.034
	(0.048)	(0.061)	(0.061)	(0.051)	(0.063)	(0.050)	(0.056)	(0.052)	(0.067)	(0.055)	(0.051)
Female Teacher	0.171***	-0.039	0.033	0.089	-0.082	-0.195***	0.017	0.138***	0.147**	-0.004	0.005
	(0.046)	(0.048)	(0.063)	(0.071)	(0.070)	(0.043)	(0.044)	(0.039)	(0.068)	(0.039)	(0.031)
Monitoring x Female Teacher		0.102	0.009	0.051	0.126		0.123*		-0.109	0.073	0.071
	(0.000)	(0.089)	(0.107)	(0.080)	(0.084)	(0.000)	(0.072)	(0.000)	(0.092)	(0.054)	(0.049)
Observations	453	453	453	452	433	432	432	427	430	2,846	3,925
R2	0.007	0.005	0.004	0.013	0.007	0.009	0.009	0.008	0.009	0.002	0.002
Adjusted R2	0.0004	-0.004	-0.005	0.004	-0.003	-0.00004	-0.002	-0.001	-0.0002	0.001	0.001
Residual Std. Error	0.481 (df = 449)	0.482 (df = 448)	0.482 (df = 448)	0.480 (df = 447)	0.487 (df = 428)	0.486 (df = 427)	0.487 (df = 426)	0.479 (df = 422)	0.485 (df = 425)	0.479 (df = 2842)	0.483 (df = 3921)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Average gratitude

	Dependent variable:
	Average gratitude
	Round 1	Round 2	Round 3	Round 4	Round 5	Round 6	Round 7	Round 8	Round 9	Pure Treatment	Perc. courses monitored
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Monitoring	-0.050	-0.080	-0.026	-0.051	-0.078	0.010	0.009	0.013	0.065	-0.001	-0.018
	(0.038)	(0.050)	(0.046)	(0.043)	(0.051)	(0.039)	(0.045)	(0.040)	(0.054)	(0.045)	(0.042)
Female Teacher	0.135***	-0.027	0.025	0.051	-0.067	-0.133***	0.059*	0.085***	0.141***	0.005	0.013
	(0.030)	(0.048)	(0.048)	(0.059)	(0.047)	(0.034)	(0.034)	(0.032)	(0.048)	(0.033)	(0.026)
Monitoring x Female Teacher		0.095	-0.023	0.051	0.123*		0.028		-0.122	0.040	0.040
	(0.000)	(0.072)	(0.098)	(0.065)	(0.064)	(0.000)	(0.055)	(0.000)	(0.075)	(0.046)	(0.041)
Observations	452	452	452	451	432	431	431	426	429	2,837	3,916
R2	0.006	0.007	0.003	0.010	0.008	0.010	0.008	0.009	0.013	0.001	0.002
Adjusted R2	-0.0002	-0.002	-0.006	0.001	-0.002	0.001	-0.003	-0.0004	0.004	0.0003	0.001
Residual Std. Error	0.392 (df = 448)	0.392 (df = 447)	0.393 (df = 447)	0.392 (df = 446)	0.398 (df = 427)	0.398 (df = 426)	0.398 (df = 425)	0.389 (df = 421)	0.398 (df = 424)	0.391 (df = 2833)	0.394 (df = 3912)
Note:	p<0.1; p<0.05; p<0.01
	Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.

Is the monitoring’s effect on IAT score due to selection bias?

Round 9 monitoring Heckman Models

monit_round9_models <- list()

for (heck.control in unlist(heck.controls)) {
  selection_formula <- formula(paste('!skip_iat ~ monitoreo  + ', heck.control))
  outcome_formula <- formula(paste('iat_score ~ monitoreo  + ', heck.control))
  
  tryCatch({
    heck.monit.round9.iat_score.cv <- selection(selection_formula, outcome_formula,
                                                data = data %>% filter(round == 9))
    monit_round9_models[[heck.control]] <- heck.monit.round9.iat_score.cv
    # print(heck.control)
    # print(summary(heck.monit.round9.iat_score.cv))
  }, error=function(e){
    # print(heck.control)
    #print(e)
  })
}

stargazer(monit_round9_models,
          type = 'html',
          dep.var.labels = c('IAT Score'),
          covariate.labels = c('Monitoring', 'Participant - Female', 'Participant - Judge', 'Participant - Years of Tenure', 'Early Responder'),
          notes=c("<span>&#42;&#42;&#42;</span>: p<0.01; <span>&#42;&#42;</span>: p<0.05; <span>&#42;</span>: p<0.1",
                  "Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables",
                  "'Early responder' started the final exercises on the same weekends as the workshop."),
          notes.append=F)

	Dependent variable:
	IAT Score
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)
Monitoring	0.019	0.019	0.017	0.886	0.607	0.018	0.397	0.008	0.913	-0.118	0.017	0.023	-0.007	0.128	-0.099
	(0.048)	(0.072)	(0.056)	(Inf.000)	(0.449)	(0.046)	(0.264)	(0.069)	(0.654)	(0.190)	(0.048)	(0.051)	(0.102)	(0.131)	(0.200)
Participant - Female	-0.015				1.731***	-0.011	1.257***				-0.011	-0.016	0.438***		0.449**
	(0.039)				(0.417)	(0.040)	(0.199)				(0.041)	(0.039)	(0.095)		(0.202)
Participant - Judge		-0.0001			1.312***			0.005	2.092***		0.004	0.008		0.388***	0.286
		(0.041)			(0.460)			(0.042)	(0.645)		(0.041)	(0.041)		(0.139)	(0.219)
Participant - Years of Tenure			-0.008			-0.007		-0.009		-0.011	-0.007		-0.124***	-0.050*	-0.033
			(0.010)			(0.009)		(0.011)		(0.024)	(0.010)		(0.022)	(0.030)	(0.040)
Early Responder				1.945			1.055***		2.409***	0.313		-0.050	0.300***	0.442***	0.289
				(Inf.000)			(0.225)		(0.615)	(0.192)		(0.039)	(0.102)	(0.124)	(0.205)
Constant	-0.177	-0.185	-0.146	-2.127	-2.514***	-0.142	-2.049***	-0.173	-3.358***	-1.116***	-0.144	-0.165	-0.882***	-0.673***	-1.282***
	(0.119)	(0.215)	(0.123)	(Inf.000)	(0.294)	(0.095)	(0.126)	(0.161)	(0.412)	(0.095)	(0.106)	(0.133)	(0.136)	(0.183)	(0.235)
Observations	370	370	370	370	370	370	370	370	370	370	370	370	370	370	370
Log Likelihood	-272.680	-272.848	-272.011	-686.564	-773.589	-271.757	-653.469	-271.939	-877.759	-555.904	-271.684	-271.784	-551.317	-445.016	-572.197
rho	0.016 (0.972)	0.033 (1.843)	0.059 (1.223)	1.000 (Inf.000)	0.990	0.035 (0.828)	1.000	0.334 (1.537)	0.990	1.000	0.054 (0.950)	0.030 (1.057)	0.996	1.000*** (0.001)	0.998
Note:	***: p<0.01; **: p<0.05; *: p<0.1
	Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables
	‘Early responder’ started the final exercises on the same weekends as the workshop.

Perc. courses Heckman Models

monit_ratio_models <- list(no_controls = heck.monit.ratio.iat_score)

for (heck.control in unlist(heck.controls)) {
  selection_formula <- formula(paste('!skip_iat ~ monitoreo  + ', heck.control))
  outcome_formula <- formula(paste('iat_score ~ monitoreo  + ', heck.control))
  tryCatch({
    heck.monit.ratio.iat_score.cv <- selection(selection_formula, outcome_formula, data = monit_perc)
    monit_ratio_models[[heck.control]] <- heck.monit.ratio.iat_score.cv
    # print(heck.control)
    # print(summary(heck.monit.ratio.iat_score.cv))
  }, error=function(e){
    # print(heck.control)
    # print(e)
  })
}



stargazer(monit_ratio_models,
          type = 'html',
          dep.var.labels = c('IAT Score'),
          covariate.labels = c('Monitoring', 'Participant - Female', 'Participant - Judge', 'Participant - Years of Tenure', 'Early Responder'),
          notes=c("<span>&#42;&#42;&#42;</span>: p<0.01; <span>&#42;&#42;</span>: p<0.05; <span>&#42;</span>: p<0.1",
                  "Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables",
                  "'Early responder' started the final exercises on the same weekends as the workshop."),
          notes.append=F)

	Dependent variable:
	IAT Score	NA
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)
Monitoring	-0.017	-0.014	0.280***	0.086***	0.232***	0.086***	0.085***	0.091***	0.084***	0.254	-0.421***	0.084***	0.091***	0.090***	-0.149***	0.091***
	(0.046)	(Inf.000)	(0.015)	(0.021)	(0.057)	(0.020)	(0.018)	(0.019)	(0.027)	(Inf.000)	(0.049)	(0.019)	(0.020)	(0.018)	(0.044)	(0.019)
Participant - Female		0.236				-0.021	-0.016	-0.020				-0.016	-0.021	-0.016		-0.017
		(Inf.000)				(0.013)	(0.013)	(0.013)				(0.014)	(0.013)	(0.013)		(0.013)
Participant - Judge			-0.176***			0.006			0.009	-0.258		0.010	0.016		0.219***	0.019
			(0.027)			(0.013)			(0.014)	(Inf.000)		(0.013)	(0.013)		(0.043)	(0.013)
Participant - Years of Tenure				-0.008***			-0.008***		-0.009**		-0.060***	-0.008***		-0.008***	-0.019**	-0.008***
				(0.003)			(0.003)		(0.003)		(0.010)	(0.003)		(0.003)	(0.008)	(0.003)
Early Responder					-0.649***			-0.064***		-0.311	0.108**		-0.066***	-0.063***	0.094**	-0.065***
					(0.045)			(0.012)		(Inf.000)	(0.046)		(0.013)	(0.012)	(0.042)	(0.013)
Constant	0.111***	-0.599	0.130***	-0.169***	0.433***	-0.207***	-0.166***	-0.184***	-0.175***	0.330	-1.187***	-0.169***	-0.188***	-0.146***	-0.716***	-0.150***
	(0.030)	(Inf.000)	(0.019)	(0.032)	(0.033)	(0.036)	(0.028)	(0.032)	(0.044)	(Inf.000)	(0.057)	(0.031)	(0.037)	(0.028)	(0.051)	(0.031)
Observations	3,390	3,390	3,390	3,385	3,390	3,390	3,385	3,390	3,385	3,390	3,385	3,385	3,390	3,385	3,385	3,385
Log Likelihood	-3,995.410	-3,420.246	-3,368.208	-2,463.315	-4,482.514	-2,468.077	-2,458.919	-2,455.920	-2,461.761	-3,949.339	-5,704.835	-2,457.232	-2,454.309	-2,446.037	-4,023.765	-2,443.654
rho	-0.998	1.000 (Inf.000)	-1.000	0.026 (0.316)	-0.999	0.015 (0.300)	0.034 (0.238)	0.015 (0.267)	0.072 (0.460)	-1.000 (Inf.000)	1.000	0.045 (0.274)	0.016 (0.305)	0.034 (0.238)	0.998	0.040 (0.269)
Note:	***: p<0.01; **: p<0.05; *: p<0.1
	Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables
	‘Early responder’ started the final exercises on the same weekends as the workshop.

Pure control monitoring Heckman Models

monit_pure_models <- list()

for (heck.control in unlist(heck.controls)) {
  selection_formula <- formula(paste('!skip_iat ~ monitoreo  + ', heck.control))
  outcome_formula <- formula(paste('iat_score ~ monitoreo  + ', heck.control))
  
  tryCatch({
    heck.monit.pure.iat_score.cv <- selection(selection_formula, outcome_formula,
                                                data = pure_control)
    monit_pure_models[[heck.control]] <- heck.monit.pure.iat_score.cv
    
     #print(heck.control)
     #print(summary(heck.monit.pure.iat_score.cv))
  }, error=function(e){
     #print(heck.control)
  #print(e)
  })
}

stargazer(monit_pure_models,
          type = 'html',
          dep.var.labels = c('IAT Score'),
          covariate.labels = c('Monitoring', 'Participant - Female', 'Participant - Judge', 'Participant - Years of Tenure', 'Early Responder'),
          notes=c("<span>&#42;&#42;&#42;</span>: p<0.01; <span>&#42;&#42;</span>: p<0.05; <span>&#42;</span>: p<0.1",
                  "Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables",
                  "'Early responder' started the final exercises on the same weekends as the workshop."),
          notes.append=F)

	Dependent variable:
	IAT Score
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)
Monitoring	-0.014	0.089***	0.086***	-0.357***	0.088***	0.086***	-0.438**	0.085***	-0.082*	0.093***	0.086***	-0.173**	0.093***	0.092***	0.092***
	(0.109)	(0.027)	(0.020)	(0.067)	(0.024)	(0.019)	(0.189)	(0.022)	(0.047)	(0.020)	(0.020)	(0.067)	(0.019)	(0.021)	(0.019)
Participant - Female	1.662***				-0.024	-0.019	1.819***				-0.018	0.443***	-0.016		-0.015
	(0.083)				(0.015)	(0.015)	(0.177)				(0.015)	(0.067)	(0.015)		(0.015)
Participant - Judge		-0.044***			-0.043***			-0.041**	0.289***		-0.040**	0.422***		-0.030*	-0.030*
		(0.017)			(0.017)			(0.017)	(0.052)		(0.017)	(0.075)		(0.017)	(0.017)
Participant - Years of Tenure			-0.017***			-0.016***		-0.017***		-0.018***	-0.016***		-0.018***	-0.018***	-0.017***
			(0.005)			(0.005)		(0.005)		(0.005)	(0.005)		(0.005)	(0.005)	(0.005)
Early Responder				1.126***			1.598***		0.209***	-0.075***		0.306***	-0.074***	-0.071***	-0.071***
				(0.053)			(0.184)		(0.049)	(0.015)		(0.070)	(0.015)	(0.015)	(0.015)
Constant	-1.450***	-0.214***	-0.141***	-1.271***	-0.205***	-0.136***	-2.319***	-0.134***	-0.706***	-0.109***	-0.129***	-1.053***	-0.105***	-0.105***	-0.101***
	(0.068)	(0.046)	(0.029)	(0.030)	(0.042)	(0.029)	(0.134)	(0.031)	(0.030)	(0.030)	(0.030)	(0.047)	(0.029)	(0.031)	(0.030)
Observations	2,432	2,432	2,432	2,432	2,432	2,432	2,432	2,432	2,432	2,432	2,432	2,432	2,432	2,432	2,432
Log Likelihood	-4,097.387	-1,792.339	-1,779.133	-3,515.098	-1,790.687	-1,777.146	-5,010.240	-1,774.010	-2,729.838	-1,766.392	-1,771.923	-3,322.238	-1,764.728	-1,762.832	-1,760.989
rho	1.000*** (0.00004)	0.024 (0.393)	0.053 (0.249)	1.000* (0.606)	0.032 (0.345)	0.044 (0.219)	1.000 (1.823)	0.074 (0.279)	1.000*** (0.153)	0.052 (0.245)	0.054 (0.234)	1.000*** (0.0001)	0.041 (0.217)	0.065 (0.263)	0.048 (0.226)
Note:	***: p<0.01; **: p<0.05; *: p<0.1
	Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables
	‘Early responder’ started the final exercises on the same weekends as the workshop.

After exploring all possible combinations of control variables: "female", "level", "judge", "years_tenure", "early_participant", on the 9th round of monitoring, the percentage f monitored courses and the pure treatment especifications we found three models with significant rhos (none of the models for the percentage of courses monitored had a significant rho):

#formula <- as.formula(paste0(outcome, ' ~ monitoreo | idcurso | 0 | aula'))
  
out_list <- list()

for (r in 1:9) {
  
  tryCatch({
    
    model <- selection(!skip_iat ~ monitoreo + female + level_2 + judge + years_tenure + early_participant, iat_score ~ monitoreo + female + level_2 + level_3 + judge + years_tenure + early_participant, data = data %>% filter(round == r))
    out_list[[sprintf('Round %i', r)]] <- model
    
  }, error=function(e){

  })
}

tryCatch({
    
    out_list[['Pure Treatment']] <- selection(!skip_iat ~ monitoreo + female + level_2 + judge + years_tenure + early_participant,
                                              iat_score ~ monitoreo + female + level_2 + judge + years_tenure + early_participant, data = pure_control)
    
  }, error=function(e){

  })

tryCatch({
    
    out_list[['Perc. courses monitored']] <- selection(!skip_iat ~ monitoreo + female + level_2 + level_3 + judge + years_tenure + early_participant,
                                                       iat_score ~ monitoreo + female + level_2 + level_3 + judge + years_tenure + early_participant, data = monit_perc)
    
  }, error=function(e){

  })

NULL

# stargazer(out_list,
#           type = 'html',
#           dep.var.labels = c('IAT Score'),
#           notes=c("<span>&#42;&#42;&#42;</span>: p<0.01; <span>&#42;&#42;</span>: p<0.05; <span>&#42;</span>: p<0.1",
#                   "Estimates are Heckman-style selection models.",
#                   "'Early responder' started the final exercises on the same weekends as the workshop."),
#           notes.append=F)

Lee (2009) approach

Lee (2009) proposes an estimator for treatment effect bounds under the assumptions of random assignment of treatment and monotonicity. We use this approach since the monitoring intervention increased the likelihood of the students skipping the IAT score sections, potentially creating bias in the IAT score estimates. We find lower bounds different from zero from round 6 onwards, suggesting an incremental effect of the monitoring intervention.

## Estimation of Lee Bounds for IAT scores

# Load Utils (leebounds functions)
source(paste0(github,"utils/utils_leebounds.R"))

# Find mean of IAT scores for those in the treatment group (monitoring)
mean_treatment_group <- data %>%
  filter(monitoreo == 1) %>%
  group_by(monitoreo) %>%
  summarize(mean_treatment = mean(iat_score, na.rm = TRUE)) 
mean_treatment_group <- mean_treatment_group[[2]]  

# Preprocess dataset for rounds
leebound_rounds.list <- data %>% 
  select("dni", "round", "monitoreo", "skip_iat", "iat_score", "female", "level_2", "level_3", "judge", "years_tenure", "early_participant") %>%
  filter(!is.na(skip_iat)) %>% 
  mutate(
    selection = !(skip_iat),
    #iat_score = case_when(
    #  iat_score == iat_score ~ iat_score,
    #  selection == TRUE  ~ mean_treatment_group), 
    years_tenure = replace(years_tenure, is.na(years_tenure), 0),
    vars_deleted = is.na(iat_score) & skip_iat == 0) %>%
  rename(treat = monitoreo, outcome = iat_score) %>%
  filter(vars_deleted == FALSE) %>% 
  group_split(round) %>% 
  as.list() %>% 
  map(pivot_wider)

# Define Control vars
control_vars <- c("female", "judge")
#control_vars <- c("female", "level_2")


# Linear model fit
leebound_rounds.list <- lapply(leebound_rounds.list, function(x){
  
  lm.fit_x<-lm(form=as.formula(paste0("outcome~",paste0(control_vars,collapse="+")) ), 
               data=x[x$selection==1,])
  
  x$predicted_iat_score<-predict(lm.fit_x,x)
  x$predicted_iat_group<-sapply(x$predicted_iat_score,Lee_grouping)
  return(x)
})

# Lee bounds -------------------------------------------------------------------

# Lee bounds without controls
estimate <- numeric()
for(i in 1:9) {                 
  out_est <- GetBounds(leebounds(leebound_rounds.list[[i]]))   
  estimate <- c(estimate, out_est) 
}

# Lee bounds without controls
estimate_r1 <-GetBounds(leebounds(leebound_rounds.list[[1]]))
estimate_r2 <-GetBounds(leebounds(leebound_rounds.list[[2]]))
estimate_r3 <-GetBounds(leebounds(leebound_rounds.list[[3]]))
estimate_r4 <-GetBounds(leebounds(leebound_rounds.list[[4]]))
estimate_r5 <-GetBounds(leebounds(leebound_rounds.list[[5]]))
estimate_r6 <-GetBounds(leebounds(leebound_rounds.list[[6]]))
estimate_r7 <-GetBounds(leebounds(leebound_rounds.list[[7]]))
estimate_r8 <-GetBounds(leebounds(leebound_rounds.list[[8]]))
estimate_r9 <-GetBounds(leebounds(leebound_rounds.list[[9]]))

# Lee bounds with controls
estimates_sharp_r1 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[1]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r2 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[2]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r3 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[3]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r4 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[4]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r5 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[5]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r6 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[6]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r7 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[7]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r8 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[8]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r9 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[9]],group_name="predicted_iat_group",treat_helps = FALSE))

# Lee Bounds for Pure control and Monit Percentage -----------------------------

## Pure control

# Preprocess dataset
leebound_pure_control <- pure_control %>% 
  select("dni", "monitoreo", "skip_iat", "iat_score", "female", "level_2", "level_3", "judge", "years_tenure", "early_participant") %>%
  filter(!is.na(skip_iat)) %>% 
  mutate(
    selection = !(skip_iat),
    #iat_score = case_when(
      #iat_score == iat_score ~ iat_score,
      #selection == TRUE  ~ mean_treatment_group), 
    years_tenure = replace(years_tenure, is.na(years_tenure), 0),
    vars_deleted = is.na(iat_score) & skip_iat == 0,
    weights = 1) %>% 
  filter(vars_deleted == FALSE) %>% 
  rename(treat = monitoreo, outcome = iat_score) 




# Linear model fit
lm.fit_pure_control <-lm(form=as.formula(paste0("outcome~",paste0(control_vars,collapse="+")) ), 
              data=leebound_pure_control[leebound_pure_control$selection==1,])

leebound_pure_control$predicted_iat_score<-predict(lm.fit_pure_control,leebound_pure_control)
leebound_pure_control$predicted_iat_group<-sapply(leebound_pure_control$predicted_iat_score,Lee_grouping)

# Lee bounds
estimate_pure_control <-GetBounds(leebounds(leebound_pure_control))

# Lee bounds with controls
estimate_sharp_pure_control <-GetBounds(Lee_sharp_bounds(leebound_pure_control,group_name="predicted_iat_group",treat_helps = FALSE))

## Monit_percentage

# Preprocess dataset
leebound_monit_perc <- monit_perc %>% 
  select("dni", "monitoreo", "skip_iat", "iat_score", "female", "level_2", "level_3", "judge", "years_tenure", "early_participant") %>%
  filter(!is.na(skip_iat)) %>% 
  mutate(
    selection = !(skip_iat),
    #iat_score = case_when(
    #  iat_score == iat_score ~ iat_score,
    #  selection == TRUE  ~ mean_treatment_group), 
    years_tenure = replace(years_tenure, is.na(years_tenure), 0),
    vars_deleted = is.na(iat_score) & skip_iat == 0,
    weights = 1) %>% 
  filter(vars_deleted == FALSE) %>% 
  rename(treat = monitoreo, outcome = iat_score) 

# Linear model fit
lm.fit_monit_perc <-lm(form=as.formula(paste0("outcome~",paste0(control_vars,collapse="+")) ), 
                         data=leebound_monit_perc[leebound_monit_perc$selection==1,])

leebound_monit_perc$predicted_iat_score<-predict(lm.fit_monit_perc,leebound_monit_perc)
leebound_monit_perc$predicted_iat_group<-sapply(leebound_monit_perc$predicted_iat_score,Lee_grouping)

# Lee bounds
estimate_monit_perc <-GetBounds(leebounds(leebound_monit_perc))

# Lee bounds with controls
estimate_sharp_monit_perc <-GetBounds(Lee_sharp_bounds(leebound_monit_perc,group_name="predicted_iat_group",treat_helps = FALSE))

# Generalized lee bounds -------------------------------------------------------
N_rep=500
ci_alpha=0.05
quantile_grid_size=0.01

# Pure control ----
form_nonmonotone<-as.formula(paste0("selection~(treat)*(", paste0(control_vars,collapse="+"),")"))

min_iat=min(leebound_pure_control$outcome[leebound_pure_control$selection==1])
max_iat=max(leebound_pure_control$outcome[leebound_pure_control$selection==1])

glm.fit<-glm(form=form_nonmonotone,data= leebound_pure_control,family="binomial")
res<-predict_selection(glm.fit, leebound_pure_control)
s.hat=data.frame(s.0.hat=res$s.0.hat, s.1.hat=res$s.1.hat)
p.0.star<-s.hat$s.0.hat/s.hat$s.1.hat
inds_helps<-(p.0.star<=1)
inds_hurts<-(!inds_helps)

# Estimates
estimates_g_pure_control<-GetBounds(leebounds_wout_monotonicity(leebound_pure_control,p.0.star))

bounds_g_pure_control<-main_bb(function_name=leebounds_wout_monotonicity,p.0.star=p.0.star,mydata=leebound_pure_control[,c("treat","selection","outcome","weights")],N_rep=N_rep)

## confidence region for identified set
CR_g_pure_control<-compute_confidence_region(bounds_g_pure_control,estimates_g_pure_control, ci_alpha=ci_alpha )


# Monit_perc ----
form_nonmonotone<-as.formula(paste0("selection~(treat)*(", paste0(control_vars,collapse="+"),")"))

min_iat=min(leebound_monit_perc$outcome[leebound_monit_perc$selection==1])
max_iat=max(leebound_monit_perc$outcome[leebound_monit_perc$selection==1])

glm.fit<-glm(form=form_nonmonotone,data= leebound_monit_perc,family="binomial")
res<-predict_selection(glm.fit, leebound_monit_perc)
s.hat=data.frame(s.0.hat=res$s.0.hat, s.1.hat=res$s.1.hat)
p.0.star<-s.hat$s.0.hat/s.hat$s.1.hat
inds_helps<-(p.0.star<=1)
inds_hurts<-(!inds_helps)

# Estimates
estimates_g_monit_perc<-GetBounds(leebounds_wout_monotonicity(leebound_monit_perc,p.0.star))

bounds_g_monit_perc<-main_bb(function_name=leebounds_wout_monotonicity,p.0.star=p.0.star,mydata=leebound_monit_perc[,c("treat","selection","outcome","weights")],N_rep=N_rep)

## confidence region for identified set
CR_g_monit_perc<-compute_confidence_region(bounds_g_monit_perc,estimates_g_pure_control, ci_alpha=ci_alpha )

# Generalized estimates
source(file.path("/Users/bran/Dropbox/WB/AMAG1/leebounds/generalized_lee.R"))

### Table with results ---------------------------------------------------------


# Table
ht <- hux(
  Round = c('Round 1', 'Round 2',  'Round 3',  'Round 4',  'Round 5',  'Round 6',  'Round 7',  'Round 8',  'Round 9',  'Pure Treatment', 'Perc. courses monitored'),
  Basic = c(estimate_r1, estimate_r2, estimate_r3, estimate_r4, estimate_r5, estimate_r6, estimate_r7, estimate_r8, estimate_r9, estimate_pure_control, estimate_monit_perc),
  Covariate_Based = c(estimates_sharp_r1, estimates_sharp_r2, estimates_sharp_r3, estimates_sharp_r4, estimates_sharp_r5, estimates_sharp_r6, estimates_sharp_r7, estimates_sharp_r8, estimates_sharp_r9, estimate_sharp_pure_control, estimate_sharp_monit_perc),
  Generalized = c(estimates_g_r1, estimates_g_r2, estimates_g_r3, estimates_g_r4, estimates_g_r5, estimates_g_r6, estimates_g_r7, estimates_g_r8, estimates_g_r9, estimates_g_pure_control, estimates_g_monit_perc),
  add_colnames = TRUE)

ht <-  add_footnote(ht, "N between 434 (round 1) and 560 (pure control) observations. Each panel reports estimated lower and upper bounds. Column (1) reports basic Lee Bounds. Column (2) reports covariate-based Lee bounds. Covariates include: female, level, judge. Column (3) reports Generalized Lee Bounds from Semenova (2021)")

ht                                        %>%
      set_bold(1, everywhere)             %>%
      set_bottom_border(1, everywhere)    %>%
      set_align(everywhere, 2, 'right')   %>%
      set_lr_padding(10)                  %>%
      set_width(0.7)                     %>%
      set_number_format(col= "Round", value = 0)

Round	Basic	Covariate_Based	Generalized
Round 1	[-0.002, 0.138]	[-0.003, 0.139]	[-0.012, 0.146]
Round 2	[0.001, 0.141]	[-0.002, 0.143]	[-0.007, 0.147]
Round 3	[-0.003, 0.143]	[-0.004, 0.144]	[-0.009, 0.148]
Round 4	[-0.002, 0.144]	[-0.003, 0.145]	[-0.008, 0.149]
Round 5	[0.003, 0.145]	[0.001, 0.15]	[0.001, 0.147]
Round 6	[0.019, 0.138]	[0.023, 0.151]	[0.019, 0.138]
Round 7	[0.022, 0.155]	[0.026, 0.168]	[0.022, 0.155]
Round 8	[0.03, 0.158]	[0.033, 0.168]	[0.03, 0.158]
Round 9	[-0.039, 0.077]	[-0.03, 0.088]	[-0.042, 0.079]
Pure Treatment	[0.001, 0.171]	[-0.004, 0.18]	[0.001, 0.171]
Perc. courses monitored	[-0.006, 0.164]	[-0.003, 0.161]	[-0.006, 0.164]
N between 434 (round 1) and 560 (pure control) observations. Each panel reports estimated lower and upper bounds. Column (1) reports basic Lee Bounds. Column (2) reports covariate-based Lee bounds. Covariates include: female, level, judge. Column (3) reports Generalized Lee Bounds from Semenova (2021)